The electroretinogram (ERG) has been the subject of intensive investigation in animals and man for many years. The relation of its major components to retinal structure is reasonably well understood and it is a useful clinical tool in the diagnosis of a small number of ocular diseases in man. Its usefulness has been limited, however, by the crudeness of the recording and stimulation techniques used in the clinical situation. It is proposed to surmount this limitation through the application to the ERG of a sophisticated non-linear systems analysis, a technique which can greatly augment the quantity of data recorded in a limited time and which utilizes a physiological stimulus rather than an intense flash to test the eye. ERGs from carefully selected patients with obvious retinal pathology and from animals with known retinal lesions will be evoked by a "white noise" stimulus and studied by means of Volterra functionals as developed by Wiener and applied to biological preparations by Marmarelis. Supplementary funds to permit evaluation of the white noise ERG in a primate animal model are requested.